Imx6q:How to caribrate cpu temp, modify thermal configuration?

cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

Imx6q:How to caribrate cpu temp, modify thermal configuration?

Jump to solution
3,657 Views
echozeng
Contributor III

Dear all,

        Env:Imx6q uboot 2009, kernel 3.0.35,android 4.2.2.

        Product envirement temp requirement: -25~70°C.

        Play video continuously, system works slowly after serveral hour .

        1, How do I know detection temp of cpu is correct?  How to calibrate cpu temp?

        2, Has any sugguestion about thermal configuration, such as TEMP_HOT, TEMP_CRITICAL, TEMP_ACTIVE?

Thanks.       

Labels (3)
0 Kudos
Reply
1 Solution
2,291 Views
lily_zhang
NXP Employee
NXP Employee

Hello, Mr Zeng:

Igor has shared 2 useful documents. I would like to add more comments for TEMP_CRITICAL, TEMP_HOT, TEMP_ACTIVE.

In default thermal.c, they are defined as the follows:

#define TEMP_CRITICAL                      373 /* 100 C*/

#define TEMP_HOT                             363 /* 90 C*/

#define TEMP_ACTIVE                                  353 /* 80 C*/

When reaching TEMP_CRITICAL, cooling device will shut down the system. When reaching TEMP_HOT, cooling device will lower CPU frequency and notify

GPU to run at a lower frequency. When the temperature drops to TEMP_ACTIVE, cooling device will release all the cooling actions. From your description, your system starts to become slower, but not shutdown, it means system may reach TEMP_HOT and starts to slow down.

Above values can be adjusted according to SoC grade and system thermal design. For example, if you are using auto grade chip, the SoC working temperature is "-40 to + 125°C". Your thermal design is also good. You can consider to increase above values as TEMP_ACTIVE (100), TEMP_HOT (110), TEMP_CRITICAL(120) etc. But please gurantee the value should not exceed max temperature supported by SoC.

View solution in original post

0 Kudos
Reply
2 Replies
2,291 Views
igorpadykov
NXP Employee
NXP Employee

Hi echo

please look at Chapter 20 Thermal Driver attached Linux Manual

and AN4539 p.46 Software Thermal Management Techniques

http://cache.freescale.com/files/32bit/doc/app_note/AN4579.pdf

Best regards

igor

-----------------------------------------------------------------------------------------------------------------------

Note: If this post answers your question, please click the Correct Answer button. Thank you!

-----------------------------------------------------------------------------------------------------------------------

0 Kudos
Reply
2,292 Views
lily_zhang
NXP Employee
NXP Employee

Hello, Mr Zeng:

Igor has shared 2 useful documents. I would like to add more comments for TEMP_CRITICAL, TEMP_HOT, TEMP_ACTIVE.

In default thermal.c, they are defined as the follows:

#define TEMP_CRITICAL                      373 /* 100 C*/

#define TEMP_HOT                             363 /* 90 C*/

#define TEMP_ACTIVE                                  353 /* 80 C*/

When reaching TEMP_CRITICAL, cooling device will shut down the system. When reaching TEMP_HOT, cooling device will lower CPU frequency and notify

GPU to run at a lower frequency. When the temperature drops to TEMP_ACTIVE, cooling device will release all the cooling actions. From your description, your system starts to become slower, but not shutdown, it means system may reach TEMP_HOT and starts to slow down.

Above values can be adjusted according to SoC grade and system thermal design. For example, if you are using auto grade chip, the SoC working temperature is "-40 to + 125°C". Your thermal design is also good. You can consider to increase above values as TEMP_ACTIVE (100), TEMP_HOT (110), TEMP_CRITICAL(120) etc. But please gurantee the value should not exceed max temperature supported by SoC.

0 Kudos
Reply